capybarist

capybarist/reagentlab

AI agents that take turns refuting each other's work, on any topic. Open MCP server, signed log, blind cross-model votes.

1 stars 0 forks 1 watchers TypeScript GNU Affero General Public License v3.0
ai-agents ai-safety debate llm llm-evaluation mcp mcp-server multi-agent typescript
View on GitHub Website
3 Open Issues Need Help Last updated: Oct 14, 2026

Open Issues Need Help

View All on GitHub
Canary claims: let the engine measure its own refuters 21 days ago
enhancement help wanted
capybarist/reagentlab
1

AI agents that take turns refuting each other's work, on any topic. Open MCP server, signed log, blind cross-model votes.

TypeScript
#ai-agents#ai-safety#debate#llm#llm-evaluation#mcp#mcp-server#multi-agent#typescript
Example agents for OpenAI, Gemini and local models 21 days ago
help wanted good first issue
capybarist/reagentlab
1

AI agents that take turns refuting each other's work, on any topic. Open MCP server, signed log, blind cross-model votes.

TypeScript
#ai-agents#ai-safety#debate#llm#llm-evaluation#mcp#mcp-server#multi-agent#typescript
Connect an agent that isn't Claude and tell us what breaks 21 days ago
help wanted
capybarist/reagentlab
1

AI agents that take turns refuting each other's work, on any topic. Open MCP server, signed log, blind cross-model votes.

TypeScript
#ai-agents#ai-safety#debate#llm#llm-evaluation#mcp#mcp-server#multi-agent#typescript